JP2002511685A - Digital watermarks and banknotes - Google Patents

Abstract

(57) [Summary] The machine-readable data is digitally watermarked on the bill (16). Such watermarks are sensed optically (13) and detected by various devices (11). In response, such devices (11) can interfere and prevent banknote duplication. This arrangement is directed to various issues, for example, using digital image editing tools to outstrip other banknote anti-copy systems.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001] This application relates to the use of digital watermarks in banknotes and other securities.

BACKGROUND OF THE INVENTION AND SUMMARY OF THE INVENTION The problem of accidental counterfeiting of banknotes first occurred twenty years ago with the introduction of color photocopying. Numerous techniques have been proposed for this problem.

[0003] US Pat. No. 5,569,628 (assigned to Ricoh) is one of several patents that can be equipped with a photocopier to recognize banknotes and prevent these photocopies. It is one of. The Ricoh patent specifically proposes that the red indicium printed on the Japanese yen is a suitable pattern for machine recognition. U.S. Pat. No. 5,845,008 (assigned to OMRON);
Nos. 24154 and 57331880 (both assigned to Canon)
FIG. 4 illustrates another photocopier that senses the presence of a seal emblem in a banknote and disables the photocopier accordingly. U.S. Pat.No. 5,678,815
No. 5 (assigned to Sharp) uses a very large number of other features on banknotes (eg, logos, stamps, seals, face value, symbols including $ and yen, etc.) as well as the basis for banknote recognition. Indicates that it can be used.

Other techniques have been proposed to prevent counterfeiting by uniquely forming the printed output from each color photocopier and allowing the copy to be traced back to the machine of origin. ing. For example, U.S. Pat. No. 5,568,268 discloses adding an essentially imperceptible pattern of yellow dots to the printed output, and U.S. Pat. No. 5,557,742 discloses a photocopier serial number output document. Discloses an associated apparatus for printing again in an essentially imperceptible format (small yellow letters). U.S. Pat. No. 5,661,574 describes that a bit containing the serial number of a photocopier, in the output printed by the photocopier, the pixel value (for example, of a yellow pixel) at a known location, and the corresponding serial number bit being "1". Shown is a device represented by increasing or decreasing by a fixed amount (eg, +/− 30) depending on whether it is “0” or “0”.

[0005] Recent advances in personal computers and color scanning and printing technology include:
Significantly improved the level of accidental forgery. High quality scanners are readily available to many computer users today, and 300 dpi scanners
Less than a dollar is available, and a 600 dpi scanner is almost available. Similarly, photographic quality color inkjet printers are generally available for less than $ 300 from Hewlett-Packard, Epson, and the like.

[0006] These tools pose new threats. For example, banknotes can be repaired (eg, by whiteout, scissors, or reclad technology), and prior art banknote detection techniques can be involved to remove / disappear visible patterns that prevent counterfeiting. Such restored documents can then be freely scanned or copied, even in photocopiers designed to prevent processing of banknote images. Adding the removed pattern back, for example by a digital image editing tool,
Free copying of the bill can be made possible.

[0007] Digital watermarking techniques provide a novel way to address the forgery problem, overcome many of the newly identified threats, and provide other benefits not previously available.

[0008] Digital watermarking (often simply "watermarking") is a rapidly evolving area of effort with several different approaches. Applicants' work has been described in U.S. Patent Nos. 5,841,978, 5,768,426, 5,748,783 and 5,748.
No. 763, No. 5,745,604, No. 5,710,834, No. 5,636,292, No. 5,721,788, and International Publication Pamphlets WO95 / 14289, WO96 /
36163 and WO 97/43736.

The applicant's International Publication Pamphlet WO 95/14289 discloses a system for redundantly embedding a plurality of bit codes in various electronic or physical media using a noise-like signal, and uses the present invention in paper in general. Indicates that it can be done. One embodiment modifies the micro-topology of the media surface and performs the encoding. Another embodiment is a scanner that embeds an identification code in the scanned output data.

[0010] Our International Publication Pamphlet WO 96/36163 discloses a duplication kiosk for use in retail photo storage devices that allows consumers to duplicate photos. The kiosk inspects consumer photos with embedded watermark data (e.g., indicating that the photo is copyrighted), and if such watermark data is detected, prevents the photo from being duplicated I do. This application describes encoding information in vector graphics and very low order bitmap graphics by slightly changing the outline of the line, for example, slightly up, down, left or right. It also discloses that it can be done.

Other digital watermarking studies are described in US Pat. Nos. 5,734,752, 56
No. 46997, No. 5659726, No. 5664018, No. 5671277, No. 5687191, No. 5687236, No. 5689587, No. 5568
No. 570, No. 5572247, No. 5574962, No. 5579124, No. 5581500, No. 5613004, No. 5629770, No. 546142
No. 6, No. 5743631, No. 5488664, No. 5530759, No. 55
No. 39735, No. 4943973, No. 5337361, No. 5404160, No. 5404377, No. 5315098, No. 5319735, No. 5337
No. 362, No. 4972471, No. 5161210, No. 5243423, No. 5091966, No. 5113437, No. 4939515, No. 534497
No. 6, No. 4855827, No. 4876617, No. 4939515, No. 49
No. 63998, No. 4969041, and International Publication Pamphlet WO 98/0286.
4. European Patent Application Publication No. 822550, International Publication Pamphlet WO97 /
39410, GB-A-2196167, EP-A-777197, EP-A-736860, EP-A-705025, EP-A-766468 , European Patent Application Publication No. 822322, International Publication Pamphlet WO95 / 2029.
1, WO96 / 26494, WO96 / 42151, WO97 / 22206, W
No. O97 / 26733. Some of the above patents relate to visible watermarking techniques. Other visible watermarking techniques (e.g., data glyphs) are described in U.S. Patent Nos. 5,706,364, 5,689,620, and 5,684,885.
No. 5,680,223, No. 5,668,636, No. 5,640,647, No. 559
No. 4809. (More typically, digital watermarks are steganographic, that is, they are essentially imperceptible to human vision and serve to hide embedded data in front of the viewer. .)

However, most of the work on watermarking is not in the patent literature, but rather in published work. In addition to the patent owners mentioned above, some other researchers in this field (text related to the watermark can be found by an author search in the INSPEC database) are described in I.S. Pitas, Eckh
ard Koch, Jian Zhao, Norishige Morimoto, Laouren
ce Boney, Kineo Matsui, A. Z. Tiekel, Fred Min
tzer, B .; Macq, Ahmed H .; Tewfick, Frederic
Includes Jordan, Naohisa Komatsu and Lawrence O'Gorman.

[0013] In a signal-theoretic method of forming a forged proof document for Szepanski's automatic verification at the Carnahan Conference in 1979 Crime Response, the author overprinted documents (eg, banknotes) with complex line patterns. Provides a deterrent to counterfeiting, but does not convey any meaningful information that can be used in automatic verification procedures. Therefore, he proposes instead to overlay a pattern representing multiple bits of data on the document. Divide the document into as many square areas as data bits to represent, and select a carrier pattern. In the area corresponding to "1" bits, the carrier pattern is added to the underlying document sample, and in the area corresponding to "0" bits, the pattern is reduced. At this time, using an automatic verification procedure, scanning the document, associating the scanned data with the carrier pattern,
Digital data can be extracted. The author specifically considers passports that process passport photographs in this way and convey the owner's name and date of birth.

The artisan is assumed to be familiar with the above prior art.

In this disclosure, it should be understood that references to watermarks not only include the assignee's watermarking technology, but may be performed in conjunction with any other watermarking technology such as those described above. is there.

[0016] Watermarks can be used for countless forms of information. This disclosure focuses in its application on banknotes, traveler's checks, passports, registered stock certificates, etc. (hereinafter collectively referred to as "security documents"), which are generally characterized by the use of line images. However, it should be recognized that the principles discussed below can be used outside of this particular area.

Most prior art in image watermarking focuses on pixelated images (eg, bitmap images, JPEG / MPEG images, VGA / SVGA displays, etc.). In most watermarking techniques, the brightness or color value of the constituent pixels is slightly changed to perform subliminal encoding of binary data through the image. (This encoding can be done directly in the pixel domain or other domain such as the DCT domain.) In some systems, the isolated pixels are changed according to one or more bits of the binary data, In other systems, it may be related to a region of pixels (eg, locally adjacent or predetermined DCTs).
Multiple populations (corresponding to components) are changed in this way. However, in all cases, the pixel functions as the ultimate carrier for the embedded data.

Although pixelated images are a relatively recent development, line art dates back several centuries. One well-known example is a US banknote. For example, in a dollar bill, line art is used in several different ways. One is to form a grid pattern surrounding the edges of the banknote (generally consisting of light lines on a dark background). Others form a grayscale image, such as a portrait of George Washington (generally consisting of dark lines on a light background).

There are two basic ways to simulate grayscale in line art.
One is to change the relative spacing of the lines to make the image area lighter or darker. FIG. 1A shows such an arrangement, where region B appears darker than region A due to the smaller spacing of the constituent lines. Another technique is to vary the width of the constituent lines, with wider lines resulting in darker areas and narrower lines resulting in lighter areas. FIG. 1B shows such an arrangement. again,
Region B appears darker than region A due to the wider configuration line. These techniques are often used together.

According to one aspect of the invention, the security of a banknote (or other security document) includes the provision of mechanically readable, generally undetectable digital data on the surface of the banknote (or other security document). Increased by marking by slightly changing the distribution of the ink. The digital data may comprise a plurality of bits and may be redundantly encoded across the bill (rather than marking the bill in only a single localized area). The encoding may use a code signal or a discrete cosine transform. In some embodiments, two watermarks may be advantageously used, in which case they may be encoded with different robustness and may be encoded according to different code signals. Some more banknotes like these also have holograms,
Security may be further improved.

According to another aspect of the invention, a banknote (or other security document) is encoded with multi-bit digital data to facilitate subsequent machine inspection.
This may be performed by receiving an initial bill copy together with the multi-bit digital data. Next, a pattern of data is generated, and the digital data is encoded in this pattern. Next, the initial composition is adjusted according to this data, and a banknote is printed using the adjusted composition. Again, the arrangement desirably provides a spatial spread of multiple bits through the bill.

One such embodiment places a virtual grid of superimposed points in a line drawing image and places the points at regular intervals in the vertical and horizontal directions. (The horizontal and vertical intervals do not need to be equal.) The virtual point may be applied to some or all of the
It may overlap at equal vertical and horizontal spacing of 250 μm. In the area of the banknote having the line drawing, the constituent lines of the line drawing meander in and between these virtual grid points.

Each grid point is considered to be the center of the corresponding area. The brightness of the area is
It is a function of the proximity of a given line to the center of the region within the boundary of the region and the thickness of the line.

In order to change the brightness of the area, the outline of the line is slightly changed in the area. In particular, the lines are made slightly thicker, less luminous or thinner, increasing luminosity. (In this example, assume a dark line on a light background.) With the ability to make these small changes, binary data is encoded in the line drawing according to known pixelation-based watermarking techniques. When such bills are subsequently scanned by a scanner, the values of the pixel data generated by the scanner reflect changes in the brightness values and allow the embedded watermark data to be decoded.

In an alternative embodiment, the thickness of the line is not changed. Instead, the position is shifted slightly towards or against the central virtual grid point, causing an increase or decrease in the brightness of the corresponding area with the same effect.

According to another aspect of the invention, duplication of a bill (or other security document) is prevented by encoding the bill with multi-bit digital data. Thereafter, sampled image data corresponding to such a bill is provided and analyzed, and the presence of the digital data there is detected. If such data is detected, take appropriate interference action.

In one such embodiment, the sampled image data is analyzed to also detect the presence of visible structural features of the bill. The interference behavior may be triggered in response to detection in the visible structure and / or digital data.

The interference includes interfering with the duplication of the image data, sending a message to a remote location to report processing of the bill data, and inserting legal tracking data into the image data. .

According to another aspect of the invention, a copy of a banknote (or other security document) is inspected for suspicious image data corresponding to a sampled optical brightness of an object, wherein the suspicious image data is Prevent by determining whether to correspond to the bill. Both of these tasks are performed at the first site. If the suspicious image data corresponds to a bill, a second remote site is contacted and a report is sent regarding the detection of bill related data.

According to another aspect of the present invention, the output from the printer of the personal computer system is marked for later identification. The method includes receiving printer data sent from the personal computer to the printer (
Or intercepting) and watermarking this data with a mechanically readable and generally imperceptible digital watermark (preferably spread through the printer data). This watermark acts to mark the printer data as being printed by a particular printer, allowing the resulting document to be combined with this printer.

According to another aspect of the invention, an apparatus for processing (eg, recognizing or confirming) banknotes (or other security documents) comprises: a scanner for generating image data corresponding to a document input to the apparatus; A processor for detecting multi-bit digital watermark data in the image data. Detection of such watermark data indicates that the document is a banknote. If the watermark data is detected, the control unit responds in a first manner and directs the operation of the device to a remote site or service (e.g., suspending duplication, inserting legal tracking information in the output data). Change or limit (by issuing an alarm, etc.). If the watermark is not detected, the control unit responds in a second different way. Scanners, photocopiers, and printers are examples of devices that can use such banknote recognition devices.

According to another aspect of the invention, a cash processing system includes an input unit for receiving a plurality of bills, and a feed mechanism for transporting the bills from the input unit past an optical detector.
The optical detector generates digital image data corresponding to the transported bill. The image data is then processed by a processor, which detects multi-bit digital watermark data that is steganographically encoded in a line drawing on at least some of the banknotes. The control unit responds in a manner that depends on the detected watermark data.

The foregoing and other features and advantages of the invention will be more readily apparent from the following detailed description, which proceeds with reference to the accompanying drawings.

DETAILED DESCRIPTION Referring to FIG. 2, an illustrative embodiment of the present invention employs a grid 10 of virtual reference points arranged on a line drawing image. The spacing between the points was 250 μm in the described arrangement
However, wider or narrower intervals can of course be used.

A surrounding area 12 shown in FIG. 3 is associated with each grid point. As described below,
The brightness (or reflectance) of each of these areas 12 is slightly changed to perform subliminal encoding of binary data.

The region 12 can take a variety of shapes, and the rounded rectangular shape shown is merely representative. In other embodiments, squares, rectangles, circles, ovals, etc. may be used instead.

FIG. 4 is an enlarged view of a portion of FIG. 3, showing a line 14 passing through the grid of points. The thickness of the line will, of course, depend on the particular image of which the line is a part. The lines shown are approximately 25 μm wide, and thicker or thinner thicknesses can of course be used.

In the first embodiment of the invention shown in FIG. 5, the width of the line is controllably changed so as to change the brightness of a region through which the line passes. Its brightness (reflectance)
The lines are made thinner (i.e., less ink in the area) to increase power. To reduce the brightness, the lines are made thicker (ie, more ink).

Whether the brightness of a given area should be increased or decreased depends on the particular watermarking algorithm used. By changing the brightness of the region 12, any algorithm can be used that changes the brightness or color of the pixels in the pixelated image in other ways. (Some watermarking algorithms make these changes in a transformed domain, such as DCT, wavelet, or Fourier. However, these changes ultimately result in changes in brightness or color. It will be clear.)

In one example algorithm, the binary data is replaced with 0s and 1s—
Expressed as columns 1 and 1. (The binary data can comprise one piece of information, but more typically some. In an illustrative embodiment, the data comprises 100 bits, some of which may be error corrected or (Detection bit.)

Next, each element of the binary data sequence is multiplied by a corresponding element of a pseudo-random sequence of -1 and 1 to produce an intermediate data signal. Each element of the intermediate data signal corresponds to an individual sub-portion of the image, such as region 12. Collectively, these elements form a pattern of binary data. (Typically, each element of the intermediate data signal is mapped to some of the regions.)

Analyzing the image (and optionally surroundings) in each such area, determining its relative capabilities, hiding the embedded data, and generating corresponding scaling factors. The example conversion factor may vary from 0 to 3. The scale factor for the region is then multiplied by an element of the intermediate data signal corresponding to (ie, mapped to) the region to produce a tweak (ie, a bias). In the case shown, the resulting tweak is between -3 and 3. Collectively, these elements form a pattern of grayscaled data.

Next, the brightness of the area on the bill is adjusted according to the corresponding tweak value. A tweak value of -3 corresponds to a -5% change in brightness, -2 is-
A -1 corresponds to a -1% change, 0 corresponds to no change, and 1 corresponds to +
A 1% change, 2 may correspond to a + 2% change, and 3 may correspond to a + 5% change. (This example follows the basic technique described in the real-time encoder disclosed in Japanese Patent No. 5710834.)

In FIG. 5, the watermarking algorithm determines that the brightness of the area A should be reduced by a certain rate and the brightness of the areas C and D should be increased by a certain rate. .

In region A, the brightness is reduced by increasing the thickness of the line. In region D, the brightness is increased by reducing the thickness of the line, and similarly in C (but to a lesser extent).

The line does not pass through area B and there is no opportunity to change the brightness of this area. However, this is because the exemplary watermarking algorithm randomly encodes each bit of data in the sub-portion spread throughout the line drawing image,
Not fatal to the method.

The changes to the line thickness in regions A and D of FIG.
Exaggerate. Although the variations shown are possible, most implementations
Typically, the thickness of the line is modulated (increased or decreased) by 3-50%.

(Many watermark processes operate in a fixed manner within a signal margin of about +/− 1% change in brightness. That is, the “noise” added by the encoding is: A line typically does not occupy the entire area of a region, and a 10% change in line thickness may only change the brightness of a region by 1%. Banknotes differ from photographs in that the artwork generally does not need to convey photorealism, thus signifying the banknotes with the greater energy used to process the photographs. Can be
The result is still aesthetically pleasing. By way of illustration, localized brightness changes on the order of 10% are possible in banknotes, and such levels of watermark energy in photographs are generally considered unacceptable. In some situations, localized brightness changes of up to 20, 30, 50 or 100% are acceptable. )

In the illustrated embodiment, the change to line thickness is a function of the tweak alone to be used for one region. Thus, if the line passes through any part of the area where a 2% tweak should be used, the thickness of the line in this area will change and 2%
Brightness difference occurs. In different embodiments, the change in line thickness is a function of the distance between the central grid point of the region and the closest approach of the line to that point. When the line passes through the grid point, a full 2% change occurs. As the distance increases continuously, successively smaller changes are made. The manner in which the magnitude of the tweak varies as a function of the position of the line in the region can be determined by using one of various interpolation algorithms such as bilinear, bicubic, cubic spline, custom curve, etc. it can.

In another different embodiment, the change in line thickness in a given region is a weighted function of the tweak with respect to an adjacent or surrounding region. In this way, the thickness of a line in an area may be increased or decreased according to the tweak value corresponding to one or more adjacent areas.

[0051] Combinations of the above embodiments can also be used.

In the embodiments described above, it is often necessary to exchange the tweak values of adjacent regions. For example, a line may pass along boundaries between regions and may pass through points (equidistant zones) equidistant from four grid points. In such cases, the line may be subject to conflicting tweak values,
Some regions may desire to increase the thickness of the line, while other regions may desire to reduce the thickness of the line. (Or, both may want to increase the thickness of the line, but may be of different amounts.) The line does not pass through the equidistant zone, but the change in line thickness changes the different tweak values. The same applies to the case where the function is in the vicinity of the possessed area. Again, by using a known interpolation function to determine the change to be made to the line thickness in a given region, the weight to be given to the tweak from each region can be determined.

In the example watermarking algorithm, the average change in brightness across the image is zero, and it is not clear that the image will be brighter or darker overall. Local changes in brightness are subtle in quantity and localized in location, so they are essentially invisible to the human eye (eg, unobtrusive / subliminal).

An alternative embodiment is shown in FIG. 6, in which the position of the line changes rather than the thickness of the line.

In FIG. 6, the original position of the line is indicated by a dotted line, and the changed position of the line is indicated by a solid line. The lines are moved slightly closer to the center of the grid point to reduce the brightness of the area, and the lines are moved slightly further to increase the brightness of the area. In this way, the line moves toward the center grid point in the area A, and moves away in the area D.

The line at the left edge of the region A indicates the normal (
Note that it does not return to the (dashed) position. This is because the area to the left of area A should also have reduced brightness. If possible, it is generally preferred not to return the lines to their normal position, but instead, keep the shifted lines shifted as they enter adjacent areas. Doing so allows for a larger net line movement within the area, and the embedded signal level is increased.

Again, the line shift in FIG. 6 has been somewhat exaggerated. A more typical line shift is 3
It is on the order of -50 μm.

One way to consider the embodiment of FIG. 6 is to use an analogy to magnetism.
The grid point at the center of each region can be considered as a magnet. This will attract or repel the line. For example, a tweak value of -3 may correspond to a strong attraction force, a tweak value of +2 may correspond to a moderate repulsion force, and so on.
In FIG. 6, the grid points in the area A indicate the attraction force (ie, a negative tweak value), and the grid points in the area D indicate the repulsive force (eg, a positive tweak value).

The analogy with magnetism is valid because the magnetic effect exerted on the line depends on the distance between the line and the grid point. Thus, a line passing near the grid point
It is more shifted in position than lines near the perimeter of the area.

(In practice, the analogy to magnetism can be much more useful than a conceptual tool. Instead, modeling the magnetic action in a computer program and analyzing the desired arrangement of the lines with respect to the grid points. Can be useful.)

Each of the variations described above in connection with FIG. 5 is equally applicable to FIG.

A combination of the embodiments of FIGS. 5 and 6 can of course be used, resulting in increased watermark energy and better signal-to-noise ratio, often with less noticeable change. Absent.

In still another embodiment, the brightness in each region is changed, leaving the lines unchanged. This can be done by interspersing small dots of ink in otherwise empty areas of the area. In high quality printing of the type used for banknotes, droplets on the order of 3 μm in diameter can be placed. (Static larger droplets exceed the perceptual threshold for most viewers.) (Regular arrays or randomly or according to a desired profile such as a normal distribution) Is easy to apply
A 1% change in brightness can be provided. (Usually, a dark dot is added to an area to reduce in brightness. An increase in brightness is caused by scattering light-colored ink, or lines that would otherwise be present in the area form a bright blank in the area. (Manufacturing reality often means that many such microdots are not printed, but are printed to some extent statistically.)

In a variant of the interspersed technique, a very thin mesh line is inserted into the line drawing and
The brightness of one or more regions can be slightly changed (so-called background line shading).

In order to decode the watermark data, the encoded line drawing image must be converted to an electronic form for analysis. This conversion is typically performed by a scanner.

Since scanners are well known, a detailed description will not be given here. Suffice it to say that a scanner conventionally uses lines of closely spaced photodetector cells that generate a signal related to the amount of light reflected from successive long rows of images. Most inexpensive consumer scanners use 300 dots per inch (dp)
With a resolution of i), the center-to-center spacing of the component photodetectors is about 84 μm. Higher quality scanners of the kind found in most professional image processing devices and photocopies are 600 dpi (42 μm), 1200 dpi (21 μm).
) Or higher resolution.

With the example of a 300 dpi scanner (photodetector spacing of 84 μm) selected, each 250 μm area 12 in the banknote approximately corresponds to a 3 × 3 array of photodetector specimens. Of course, and only in rare cases, a given area is physically recorded by the scanner such that the nine photodetector specimens gain brightness in this area and not others. More generally, the linework is oblique with respect to the photodetectors of the scanner and is not longitudinally aligned (ie, some photodetectors image sub-portions of two adjacent regions). However, since the scanner oversamples the regions, the brightness of each region can be determined unambiguously.

In one embodiment, the data scanned from the line drawing is collected in a two-dimensional array,
Processing according to one of the techniques disclosed in our earlier patents and applications, analyzing the statistics of the data (using the techniques disclosed in our earlier documents) and Extract the data bits.

(Again, reference to our earlier watermark decoding techniques is by way of example only. Starting with scanning, once the data is available in pixel format, what other watermark decoding techniques are available. And extracting the correspondingly encoded watermark is straightforward.)

In a different embodiment, the scanned data is not collected in a complete array prior to the processing. Instead, it processes it in real time, as it was generated, to detect the embedded watermark data without delay. (Depending on the parameters of the scanner, the line drawing image, on the order of half an inch, may need to be scanned before the statistics of the resulting data clearly indicate the presence of a watermark.)

According to another aspect of the invention, various hardware devices have the ability to recognize embedded watermark data in any line drawing image that these devices process and respond accordingly. Provided.

One example is a color photocopier. Devices such as these use a color scanner to generate sampled (pixel) data corresponding to an input medium (eg, a dollar bill). If watermark data relating to a bill is detected, the photocopier performs one or more steps.

One option is simply to abort the duplication and display a message reminding the operator that the current duplication is illegal.

Another option is to connect to a remote service or site and report a copy of the bill being attempted. Photocopiers with telephone dialing capabilities
It is known in the art (eg, US Pat. No. 5,305,199) and is readily adapted for this purpose. Similarly, a personal computer having a modem or network communication capability can send a message to a remote location (eg, on the Internet) to change the copy being attempted. The remote service may be an independent service or may be a government agency.

Yet another option is to allow the duplication, but insert legal tracking data into the resulting duplication. This tracking data can take various forms. Steganographically encoded binary data is one example, and the data embedding described in US Pat. No. 5,568,268 may be used. The tracking data may record the serial number of the machine that made the copy and / or the date and time when the copy was made. Such tracking data is not normally inserted into the photocopied output to communicate personality, but only if the photocopied one is detected as a banknote.
(Such an apparatus is shown in FIG. 7).

Preferably, the scan data is analyzed on a line-by-line basis to detect an illegal photocopying process with the shortest delay. When a bill is scanned, one or more lines of scanner output may be provided to the photocopying machine's duplication technology unit before a bill detection decision is made. In this case, the photocopy has two areas, a first area without tracking marks and a second resulting area with tracking data inserted.

Another hardware device that can use the principles described above is a stand-alone scanner. A programmed processor (or dedicated hardware) inside the scanner analyzes the data generated by the device and responds accordingly.

Yet another hardware device that can use the principles described above is a printer. A processor internal to the device analyzes the graphical data to be printed and looks for watermarks related to the banknote.

For both the scanner and the printer device, the response method can include disabling operation or inserting tracking information. (Such devices typically do not have dialing capabilities or network connections, but if desired, these capabilities of conventional computers can be incorporated into scanner or printer devices using known techniques.) .)

Again, it is desirable to process the scanner or printer data as it becomes available so that any bill processing is detected with the shortest delay. Again, there is some delay before making the detection decision. Thus, the scanner or printer output consists of two parts, the part without the tracking data and the other part with the tracking data.

In another embodiment (FIG. 10), a non-perceptible watermark with a universal ID (UID) is inserted into every document printed by a printer, scanned by a scanner, or copied by a photocopier. . The UID relates to a particular printer / photocopier / scanner in a registration database maintained by the product manufacturer. The manufacturer, in this database,
It is also possible to enter the name of the seller who originally shipped the product. Further, the owner's name and address can be added to the database when the product is registered for warranty service. While not preventing the use of such machines in counterfeiting,
The embedded UID facilitates identifying opportunities for generating counterfeit banknotes. (This is an application where personal watermarking is best used.)

Another advantageous use of digital watermarks is in the reliable authentication of banknotes, for example, for automatic teller machines that both receive and pay cash. Referring to FIG. 8, such a machine (11) is provided with a known optical scanner (1).
3) to generate digital data corresponding to the surface of the bill (16). Next, the image set (14) is analyzed to extract embedded watermark data. In watermarking techniques that require knowledge of the code signal (20) for decryption (e.g., noise modulated signals, secret keys, spread signals, etc.), banknotes may be watermarked according to some such codes. Good. Some of these codes have publicly allowed them to be read by conventional machines. Other codes are
It is confidential and restricted for use by government agencies. (See: Public and private codes in patents granted by the applicant.)

As noted, banknotes may now have several visible structures or markings (eg, shown in the patents cited above) that can be used to help indicate authentication (by visual inspection or machine detection). Seal). Preferably, the bill is inspected by an independent detection system (24) to confirm the presence of both the visual structure (22) and the digitally watermarked data before determining that the bill is genuine. I do.

The visible banknote structure can be sensed using known pattern recognition techniques.
Examples of such techniques are described in U.S. Patent Nos. 5,321,773 and 5,390,902.
No. 59, No. 5,533,144, No. 5,538,841, No. 558,614, No. 5
No. 6,339,952, No. 4,723,149 and No. 5,424,807 and published foreign application EP-A-766449.

Referring to FIG. 9, for example, a photocopier (30) senses the presence of either the visible structure (32) or the embedded bill watermark data (34), and if either is present, duplicates Can be disabled. Scanners and printers can be equipped with a similar ability to analyze scanned or printed data for any of these bill verification imprints. If either is detected, software (or hardware) disables further operation.

Validation of banknotes with watermark data offers important advantages that are otherwise unavailable. As shown, the original note can be tampered with (eg, by whiteout, scissors, or rescue techniques) to remove the visible structure. The document is then converted to a visible structure sensitive photocopier or scanner /
It can be freely copied on any of the printer devices. Next, the removed visible structure can be added by a second printing / photocopying operation. If the printer is not equipped with a banknote disabling capability, an image editing tool is used to insert the visible structure back into the image data set scanned from such a tampered banknote and print the full banknote freely. can do. By additionally including and sensing watermark data embedded in the banknote, such a strategy does not succeed.

(A similar strategy is to scan the banknote image in a non-banknote sensing scanner. Then, edit the resulting data set with conventional image editing tools to remove the visible structure. Such a data set can then also be printed on a printer / photocopier which checks such data for the presence of a visible structure. / Can be inserted by a photocopy processing operation.

Preferably, the visible structure detector and the watermark detector are integrated as one hardware and / or software tool. This device has various economic benefits, such as interfacing with scanners, manipulating pixel data sets for pattern recognition and watermark extraction, electronically re-aligning images, and facilitating pattern recognition / watermark detection. , A control signal (e.g., a disabling signal) to the photocopier / scanner.

While the above-described application has disabled potential counterfeiting operations in response to detecting either a visible structure or watermarked data, in other applications both criteria have been set Needs to be satisfied before it can be recognized as genuine. Applications such as these typically include, for example, the receipt or acceptance of banknotes by the ATM described above and shown in FIG.

(To provide a comprehensive disclosure without unduly lengthening the following specification, Applicants include by reference the patent documents cited above.)

From the above, it will be appreciated that embodiments of the present invention allow a line drawing image to serve as a subliminal carrier for binary data. In addition, the deterrents that exist against banknote counterfeiting have been emphasized, preventing the general working environment and providing other benefits.

Having described and illustrated the principles of the present invention with respect to several illustrative embodiments, recognize that these embodiments are illustrative only and should not be taken as limiting the scope of the invention. Will be done. Obviously, guided by the techniques described above, other watermarking, decoding and anti-counterfeiting techniques can be used in place of and / or in combination with the detailed elements to produce similar effects.

For example, while the invention has been described with reference to an embodiment that uses a regular rectangular array of grid points, one skilled in the art can substitute other arrays of points that are neither rectangular nor regular. Will recognize.

Similarly, while the invention has been particularly described with reference to banknotes, these principles are equally applicable to other security documents.

Although the invention has been particularly described with reference to an embodiment in which the embedded energy is measured according to local image features, in other embodiments a manually generated energy profile can be realized. That is, a mask that manually defines the magnitude of the embedded signal in different parts of the image can be devised and used to create a change in brightness in each region.

Similarly, one encoding technique has been specifically described, but the invention is not so limited.
Any technique for hiding multi-bit binary data that can be detected from the sampled optical data corresponding to the security document in the security document can be used as well. Some but not all of these techniques involve minor changes to the ink (eg, color, density, distribution, etc.). For example, another approach is to digitally watermark the underlying media (paper, polymer, etc.). This can be done by changing the microstructure of the medium (small Braille) and revealing the pattern of the encoded data. Such texture processing can be optically detected and decoded using an algorithm corresponding to the encoding algorithm that generated the patterned data. Another option is to use the brightness on or in the banknote, where the brightness has the watermarking specified on / in it. The brightness can be textured (as described above) or its optical clarity can be varied as well. Approaches such as these are detailed, for example, in Applicants' International Publication Pamphlet cited above.

Although the present invention has been particularly described with reference to some hardware devices (eg, scanners, photocopiers and printers) and some applications (forgery prevention and banknote authentication), the technology is not limited thereto. Not done. For example, many different devices may be provided with the bill recognition capabilities detailed herein. One example is a cash dispenser that checks bills for authentication. Another example is a personal computer that examines the data processed thereby, determines if it is of the same origin with respect to the banknote, and intervenes appropriately.

[0098] Still other applications can be handled similarly by the individual technology. One is banknote processing, for example, a device used by banks to count, sort, read, and select banknotes that have passed a certain number of years from circulation and counterfeit banknotes. Such an apparatus may include an input (eg, a hopper) for receiving a plurality of bills, and a feed machine for transporting the bills from the input through an optical detector. The optical detector generates digital image data corresponding to the transported bill. The image data is then processed by a processor, which detects steganographically encoded multi-bit digital watermark data in the artwork for at least some of the bill. The control unit responds in a manner responsive to the detected watermark data. For example, the control unit may cause the device to sort bills by denomination or age, or select forgery, based on the decrypted data.

Furthermore, the methods of the present invention can be performed by hardware, software, or a combination (eg, performing some image processing operations with dedicated hardware and performing other operations with software). Will recognize.

[0100] In yet another embodiment, at least a portion of the watermark is printed using luminescent ink. This allows, for example, a merchant who is presented with a note to quickly recognize the presence of some watermark-like mark on the note (eg, by inspecting under a black light) without resorting to scanners and computer analysis. To be able to Such luminescent inks can also print human readable indicia on the bill, such as the face value of the bill. (Because ink jet printers and many other common printing techniques use cyan / magenta / yellow / black to form colors, they can only generate a limited spectrum of colors. Emitting colors such as yellow, pink and green dyes used for highlighting marks can be used as well, with the advantage of being visible without black light.

An improvement over the coding techniques detailed above and other coding techniques is to add an iterative assessment of the robustness of the digital watermark and a corresponding adjustment in the re-watermarking operation. In particular, when encoding a multi-bit watermark, the underlying content features (eg, artwork) may result in some bits being encoded more robustly (eg, stronger) than others. In an illustrative technique used for this improvement, the watermark is encoded first. Next, a test decoding operation is performed. Next, a reliability measurement on each bit decoded in the decoding operation (eg,
Evaluate the signal-to-noise ratio). Identify the bits that appear to be weakly encoded and make corresponding changes to the watermarking parameters to increase the relative strength of these bits. Next, the object is newly watermarked with the changed parameters. This process is required until all bits making up the encoded data are approximately equally detectable from the encoded object, or
Iteration can be performed until some predetermined signal to noise ratio threshold is met.

Banknotes, as well as most other media and objects, generally benefit from the use of multiple watermarks. For example, once a bill is marked in a spatial area,
The second time it is possible to mark in the spatial frequency domain. (It should be recognized that any change in one area has an effect in another area, where reference is made to the area where the change is directly affected.)

Another option is to use a note (or other physical or digital object)
Mark watermarks of two different levels of robustness or strength. More robust watermarks are resistant to various forms of tampering and are detectable in the object even after multiple occurrences of interfering distortion. Less robust watermarks can be formed that are weak enough to ruin the initial distortion of the object. In banknotes, for example, less robust watermarks serve as authentication marks. Any scanning and reprinting operations make this unreadable. Both robust and weak watermarks should be present in genuine banknotes, only the former being present in counterfeiting.

Yet another option is to encode two different watermarks according to two different code signals. One watermark may identify (eg, by a serial number) the device used in the attempted counterfeit operation. The other watermark can be used for other purposes.

The above discussion has directed various technical states to various purposes, and has been particularly described with respect to banknotes. An example solution has been detailed. Others will be apparent to those asking them by extrapolation from the solutions given above using their usual knowledge.

For example, the techniques and solutions disclosed herein use elements and techniques from the cited references. Other elements and techniques from the cited documents may simply be combined to yield other implementations within the scope of the invention. in this way,
For example, holograms with or without watermark data can be used in banknotes for additional security, single-bit watermarking can generally replace multi-bit watermarking, and use imperceptible The techniques described and described can often be performed using visible watermarks (such as pictograms), providing a local measure of the watermark energy and increasing the signal-to-noise ratio of the watermark without increasing human perceptibility. And various filtering operations can be used for the various functions, and the decoding can proceed at a granularity of one pixel (or DCT coefficient) or of the neighboring pixels (or DCT coefficients). Groups may be treated similarly and the encoding may be optimized to withstand the expected form of content tampering. (All cited variants are generally detailed in the applicant's International Publication Pamphlet.) Thus, the exemplary embodiment incorporates the teachings referred to above. It is just a selected sample of the solutions available by combining. Other solutions are not necessarily exhaustively described herein, but are within the full understanding of those skilled in the art given the above disclosure and familiar with the cited art.

Miscellaneous (Most of the description below was originally given in a US priority application prior to the above discussion of banknotes. This claim emphasizes the use of banknotes Therefore, the following subject matter has been moved to the end of the specification for purposes of clarity, and although not specifically described with reference to banknote applications, this subject material may provide a more complete understanding of the banknote application. Give useful contents that can be.)

An improvement to existing watermark coding techniques is to add an iterative assessment of the robustness of the mark and a corresponding adjustment in the re-watermarking operation. In particular, when encoding a multi-bit watermark, the features of the underlying content may result in some bits being more robustly (eg, more strongly) encoded than others. In an illustrative technique used for this improvement, the watermark is encoded first. Next, a test decoding operation is performed. Next, a reliability measurement (eg, signal-to-noise ratio) for each bit decoded in the decoding operation is evaluated. Identify the bits that appear to be weakly encoded and make corresponding changes to the watermarking parameters to increase the relative strength of these bits. Next, the object is newly watermarked with the changed parameters. This process is required until all the bits making up the encoded data are approximately equally detectable from the encoded object, or meet some predetermined signal to noise ratio threshold Until it can be repeated.

Although the above analysis evaluated reliability on a bit-by-bit basis, the iterative procedure involved can evaluate reliability on a part-by-portion basis. That is, consider the encoded objects in parts and analyze each part with respect to the robustness of the data encoded thereby. In those portions that indicate "weak" coding, the coding parameters can be adjusted to increase the coding in one or more subsequent recoding operations.

These parts can take different forms, for example rectangular patches in still or moving images, short temporal exclusions in audio or video, specific in the coefficient-based representation of the object in the transformed domain DCT / Fourier / Weblet coefficients (or adjust groups of coefficients) can be taken.

With this technique, even if the encoded object is spatially or temporally extracted or filtered (eg, spectrally), the watermark energy remaining after such processing enables accurate decoding. Guarantee is increased.

In an illustrative embodiment, this iterative process is highly automated and essentially transparent to the user. The user simply instructs the computer control system to watermark the object, which test watermarks, decodes, makes continuous adjustments, and produces a final encoded object that meets the desired watermark quality requirements. Respond by repeating until it occurs.

[0113] Watermarking can be used for a large number of information formats. These include digital formats (e.g., pixel images, digital video, MP3 / MP4
Includes images (including video) and audio expressed either in audio representations or analog representations (eg, unsampled music, printed images, banknotes, etc.). Watermarking can be used on digital content (eg, images, audio) before or after compression. Watermark processing is performed by Structured Au
dio, Csound, NetSound, SNHC Audio, etc. (http: // s
various "descriptions" or "synthesis" of content such as sound.media.mit.edu/mpeg4/)
It can also be used in linguistic expressions by specifying watermark data and a synthesis command that produces the intended audio signal. Examples include paper, plastic, laminates, paper / film emulsions, etc. The watermark can embed a single bit of information or any number of bits.

The physical representation of the watermarked information is in the form of changed signal values, such as slightly changed pixel values, image brightness, image color, DCT coefficients, instantaneous audio amplitude, etc. Is most commonly taken. However, watermarks can be used to create other changes, such as changes in the surface microstructure of the medium, localized chemical changes (eg, in photographic emulsions), localized changes in optical density, localized changes in luminescence. It can also be shown in the method. The watermark can optionally be implemented in holograms and conventional paper watermarks.

[0115] One improvement to the existing technology is the established web crawler (Crawle).
r) Services (eg, AltaVista, Excite or Inktomi)
), And in addition to these normal data gathering / indexing operations, watermarked content (on the web, Internet newsgroups, BBS
Search for watermarked content (in systems, online systems, etc.). Crawler such as these may have files (eg, * .JPG, * .WAV, etc.) that may have a watermark embedded for later analysis.
Can be downloaded. These files, as explained below,
Can be processed in real time. More generally, files such as these are queued and processed by a computer different from the crawler computer. Instead of performing a watermark reading operation on each such file, screening techniques can be used to recognize those most likely to be carrying the watermark data. One such technique is to perform a DCT operation on the image and look for spectral coefficients for a particular watermarking technique (eg, coefficients for a tilted and embedded subliminal grid). To decode a watermark based on a spread spectrum, the analyzing computer requires access to the noise signal used to spread the data signal.
In one embodiment, interested parties provide these noise / key signals to the crawler service to enable localization of these marked content. The crawler service keeps this information confidential and watermarks the different noise information into the decoding of certain images (where the images were used as a convenient shortening for images, video and audio). Use until data (if any) is found. This is the use of web crawlers,
Instead of just a publicly encoded watermark, it allows to locate content with a personally encoded watermark. Queuing content data for analysis offers some opportunities for computational shortcuts. For example,
Images of equal size (eg, 256 × 256 pixels) are tiled into a larger image and inspected as a unit for the presence of watermark data. When the decoding technique (or any pre-screening technique) uses DCT transform or the like, the block size of the transform can be tiled to correspond to the tile size (or an integral fraction thereof). Next, the blocks indicated as having a watermark are subjected to a complete read operation. If the queued data is classified by file name, file size or checksum, duplicate files can be identified. Upon confirming such duplication, the analysis computer:
Only one example of the file needs to be considered. If watermark data is detected from such a file, the content provider can be notified of each URL that found a copy of the file.

Some commentators have recognized that web crawler-based searches for watermarked images can be broken by breaking the watermarked image into sub-blocks (tiles). HTML instructions and the like cause the sub-blocks to be generated in a tiled fashion, recreating a complete image.
However, with small sized component sub-blocks, watermark reading is not easily performed.

This attack is overcome by instructing the web crawler to gather display instructions (eg, HTML) that position the image file relative to the display on the web page, in addition to the image failures themselves. Prior to scrutinizing files collected from certain web pages for watermarks, they can be aligned in the arrangement specified by the display instructions. With this arrangement, the tiles can be reassembled and the watermark data can be easily reproduced.

Another such hypothetical attack on web crawler detection of image watermarks is to scramble the images (and thus the watermarks) in the file and to use Java
a Unscramble before viewing the image using an applet or the like.
Existing web crawlers inspect the files as they find them, and consequently do not detect watermarks. However, just using the same applet on the web crawler as well, so as to allow the Java descrambled applet to be invoked if the user wants to access the file, overcomes such attempted tricks of watermark detection. be able to.

Even though the location of “content” can be located and indexed by various web crawlers, the content of the “content” is unknown. For example, *. JP
The G-file may contain pornography, sunset photos, etc.

Using a watermark to permanently erase metadata in content (as opposed to being stored in a data structure that forms the other part of the object, conventionally done with metadata) Can be associated. The watermark is "sunset"
And the like. Still further, the watermark may include (or consist entirely of) a unique ID (UID) that serves as an index (key) to a networked remote database that includes the metadata descriptor. Remote data may include metadata described by an extensible markup language (XML) tag set. With such an arrangement, a web crawler or the like can extract and index the metadata descriptor tags, making the search based on a semantic description of the file content rather than by file name. Enable.

[0121] Existing watermarks generally embed information that helps convey copyright information.
Some systems embed text indicating the copyright owner. Other systems embed a UID that is used as an index into a database that stores the copyright user's name and related information.

Given the future, watermarks should be more useful than silent copyright warnings. One option is to use watermarks and embed “intelligence” in the content. One form of intelligence knows its "home". “Home” may be the URL of a site to which the content relates. For example, a picture of a car can be watermarked with data identifying the website of the car dealer that issued the image. Wherever the image goes, this serves as a link back to the original distributor. The same technology can be used for the company logo. Wherever they are copied on the Internet, a suitably equipped browser or the like can decrypt the data and link back to the company home page. (Decryption, position the cursor on the logo,
This can be done by pressing the right mouse button, the pressing of the right mouse opening a window of choices, one of which is the decoding of the watermark. )

In order to reduce the data load of the watermark, it is not necessary to encode the intelligence all in the watermark of the content. Alternatively, the watermark may then again provide a UID identifying a remote database record from which the car dealer's URL or the like can be retrieved. In this way, the images and the like become marketing agents, linking the customer to the seller (with the same visual sales skills deployed). Unlike the copyright example, where distribution of images is an evil that requires tracking and stopping, distribution of the images can be treated here as an opportunity to sell. The watermarked image is the gateway to commercial transactions.

(Using an intermediate database between the watermarked content file and its final home (ie, indirectly linking) provides a significant advantage, which is Allows one to easily change the "home" by updating records in the database, such as when one company is acquired by another company, The smart image can be directed to the new company's home web page by updating the database record, whereas the old company's home URL is hard-coded (eg, watermarked) in the object. If so, this may be directed to a URL that is eventually discarded. , Serving as a switchboard coupling the file to its current home.

The above technique is not limited to digital content files. The same approach is equally applicable to printed images and the like. For example, a printed catalog may include photographs describing a jacket. The watermarked data is embedded in this photo. This data can be extracted by a simple hand scanner / decoder device using simple manipulation (eg, known to those skilled in the art) and decoding techniques. In watermark reading applications, such as with hand scanners, it is important to make the watermark decoder robust to image rotation, since the catalog photograph is likely to be scanned off-axis. One option is to encode a subliminal scale (eg, a visual synchronization code) in the catalog photo,
The set of image data is pre-processed to restore proper alignment before decoding.

The scanner / printer device may be a computer with a modem, a telephone, or
It can be coupled to some other communication device. In the former case, the device comprises:
The URL data linking the browser to the catalog seller's order page is provided to the computer's web browser. (The device does not need to include its own watermark decoder, and this task can be performed by the computer.) The seller's order page, as is routinely performed by online merchants, uses the jacket The size and color selectors can be detailed and order instructions (credit card numbers, delivery selectors, etc.) can be requested. Such a device connected to a telephone can dial the catalog seller's free automatic order telephone number (known, for example, from the data encoded in the watermark) and confirm the jacket to an order center. . The voice prompt may then request the selection of the customer's size, color, delivery selector, which may be touch-tone instructions or pronounced words (using known voice recognition software at the seller's facility). To enter.

In such applications, the watermark may be conceptualized as an invisible barcode used in purchase transactions. Here, the watermark serves as a seamless interface bridging the print and digital world.

Another way to add intelligence to content is to use watermarks and Java
a or ActiveX code. The code can be embedded in the content or stored at a remote location and linked to the content. Upon activation of the watermarked object, the code can be executed (automatically or at the user's discretion). This code can perform virtually any function. One is "call home", which launches a browser and links to the home of the object. The object is
Any data can be relayed to that home. This data may specify some attributes of the data and its use. The code may also prevent access to the underlying content until permission is granted. One example is a digital movie, which, when double-clicked, automatically runs a watermarked Java applet that links to the movie's distributor via a browser. Next, the user is prompted for a credit card number. After the number is verified and billed, the applet passes the contents of the file to a browser on the computer for viewing the movie. Support for these operations is provided, preferably via the computer's operating system or plug-in software.

If the operating system knows the content format (eg, in an object-oriented file system or “media aware”
(In the file system), giving the operating system similar functionality,
Special processing of the watermarked content can be triggered. One example application is in the acquisition of content from external audio / image / video devices.

Using such devices, it is also possible to accumulate user-provided demographic information when smart image content is accessed by a consumer of said content. The demographic information can be written to a remote database and used for market research, customizing or personalizing information about the content provided to consumers, selling opportunities, advertising, and the like.

In audio and video, watermarks include links to WWW fan sites, biographies of actors, tying (T-shirts, CDs, concert tickets)
Useful for carrying relevant information, such as advertising sales. In such applications, it is desirable (but not essential) to display a small logo and indicate the presence of additional information on the user interface (eg, screen). If the consumer selects the logo via a selection device (mouse, remote control button, etc.), the information is shown to the consumer, where he can interact with it.

Much has been written on the topic of managing asset rights (patent granted)
. Sample patent documents are disclosed in U.S. Patent Nos. 5,715,403 and 5,683,844.
3, No. 5,634,012 and 5,629,980. Again, much of the technical work has been documented in journal articles, and these are referred to IBM's
The system can be recognized by searching for relevant company names and trademarks, such as Portland Software's ZipLock system, Rights Exchange Services by SoftBank Net Solutions, and the DigiBox system by Internet Trust Technologies.

The example asset management system forms available content in an encrypted form (eg, from a web server or on a new computer's hard disk). Data that identifies the content (eg, a preview) and data that specifies various assets related to the content are associated with the encrypted content. If the user desires more complete use of the content, the user gives a billing authorization (credit card) to the distributor, who then gives a decryption key and grants access to the content. (Systems such as these are often implemented using object-based technology. In these systems, the content is commonly referred to as being distributed in "secure containers.")

Other asset management systems distribute in unencrypted “clear” or in secure containers. Devices such as these are common, for example, in enterprise asset management systems (eg, in contrast to e-commerce applications where container-based approaches are prevalent).

Preferably, the content should be marked (personalized / serialized) so that any illegal use of the content can be tracked. This marking can be done by a watermarking process, which ensures that the marks move together in whatever form and where the content goes. The watermarking of the (encrypted) object, such as by encoding the relevant UID in a database in a specific container, can be performed by the distributor before distribution. Given access to the content (eg, granting access rights to a secure container or making unencrypted content available to a user), update the database record and /
Reflect the user, purchase date, granted rights, etc. An alternative is to include a watermark encoder in the software tool used to access the content (eg, a decryption engine for a secure container-based system).
. Such an encoder may embed watermark data in the content before it is provided to the user (eg, in a secure container system, before removal from the container). The embedded data may include a UID as described above. This UID can be assigned by the distributor before distributing the content / container. Instead, the UID is
It can be a data string that is not known or formed until access (or access rights) is granted. In addition to the UID, the watermark may include other data unknown to the distributor, for example, information specifying when and how to access the content.

In still other non-container based systems, access rights are again
This can be realized by a watermark. For example, full resolution images can be freely available on the web. If a user wants to incorporate the image into a web page or magazine, the user can ask the image about the duration and status of its use. This may result in a link (directly or via an intermediate database) to the website specified by the embedded watermark, which link specifies the desired information. The user can then arrange the required payment and use the image knowing that the required rights are protected.

Image / video / audio tagging indicates that at the time of distribution to these end users (eg, from a managed content warehouse), it does not need to consist exclusively of UIDs. Other data, such as serial numbers, recipient identities, distribution dates, etc., can also be watermarked into the content. In addition, the watermarked data can help establish a permanent link to metadata contained in an associated asset management database.

As shown, digital watermarking can also be implemented using conventional (eg, paper) watermarking techniques. Known techniques for watermarking media (e.g., paper, plastic, polymer) are described in U.S. Patent Nos. 5,536,468 and 52,758.
Nos. 70, 4760239, 4256652, and 4370200, which can be adapted to display a visual watermark instead of a logo or the like. Note that some forms of conventional watermarks designed to be seen by transmitted light also appear as low-level signals in reflected light as typically used in scanners. Using a transmitted illumination detection system, such watermarks can also be detected using conventional optoelectronic watermark detection techniques known in the art.

As also shown, the digital watermark can be realized as part of an optical hologram. Known techniques for generating and safely mounting holograms are described in U.S. Patent Nos. 5,319,475, 5,694,229, 5,492,370, and 548.
Nos. 3,363, 5,658,411 and 5,310,222. To watermark a hologram, the watermark can be represented in an image or data model that produces a holographic diffraction grating. In one embodiment, the hologram is generated as usual, displaying objects or symbols. The watermark markings appear in the background of the image, so that they can be detected from all viewing angles. In this context, it is not necessary that the watermark representation be essentially imperceptible to the viewer. If desired, a clearly visible noise-like pattern can be used without compromising the application in which the hologram is placed.

Digital watermarks can also be used with labels and tags. In addition to the conventional label / tag printing process, other security-equipped techniques can be used. Known techniques useful in generating security labels / tags are described in U.S. Patent Nos. 5,665,194, 5,732,979, and 5651.
Nos. 615 and 4,268,983. The imperceptibility of watermarked data and the ease of machine decoding are some of the benefits associated with watermarked tags / labels. In addition, the cost is significantly lower than many related technologies (eg, holograms). Using the watermark in this application, the authenticity of the product label can be assured to either the merchant or the customer of the relevant product using a simple scanner device, whereby the sale of counterfeit products Decrease the ratio.

Conveniently, ignoring with respect to time, which is an inevitable social ramification of the following devices, considers a steganographic setup, whereby certain computing systems based on the “master clock” drive function require a small Alternatively, the computing system may add a signal in response to a uniquely identifiable electrical operation when the system communicates digitally with a second computing system. The minor changes involved have no effect on the basic operation of the first computing system,
It does not affect communication between the two systems.

Social branching involves referring to ideas where purely anonymous concepts in Internet content still exist and related ideas of the right to privacy. It seems that the "good thing" of this concept is that non-anonymity may be a source of explosion, and the "bad thing" of this concept escapes the illusion of responsiveness to one's own actions Seems to be that. All may be the case, and seems to be the basis for a socially neutral concept of simply identifying a given physical object in comparison to its neighbors, and therefore, The approach provides a unique signal to the computing system so that a second system communicating with the first system can identify the first system.

Using the spread spectrum principle (eg, modulating a “noise-like” carrier signal with an information signal to produce a data signal), modulating the instantaneous phase of the computer clock signal and encoding it steganographically Transport the data. In one embodiment, the phase of the clock frequency changes instantaneously according to the data signal. (In some embodiments, the data signal need not reflect modulation by a noise-like carrier signal and can carry an information signal that does not change.) In an illustrative system, the period of the clock signal is 3 nanometers. Seconds
The instantaneous phase shift is on the order of +/- 10 picoseconds. (These numbers will, of course, depend on the particular microprocessor used and its tolerance for phase noise when still running within the specification.) If a binary "1" is to be encoded, the phase will be 10 picoseconds. Seconds. If a binary "0" is to be encoded, the phase is delayed by 10 picoseconds. Successive clock edges are continuously advanced or delayed according to the corresponding bit of the data signal until the end of the data is reached. Next, the data signal is desirably repeated to provide redundant coding.

Such modulation of the computer's clock system is manifested in its operation and includes its communication with connected devices. Thus, for example, a modem or network connection carries data affected by the phase modulation of the clock signal. In particular, any such communication (eg, RS-232C, TCP / I
P, wireless modem communications, etc.) does not occur at evenly spaced clock edges, but instead, its precise temporal characteristics vary slightly according to the data signal that modulates the clock signal. Occurs at the edge.

A remote computer that receives the information sent from such a computer can examine the timing of the signal (eg, changes in its edges) and thereby identify the encoded data signal.

If the channel to the remote computer provides near-perfect temporal fidelity (ie, does not inherently introduce a variable delay into the propagation of the data), the data may be interpreted by its prediction of each edge change. Can be extracted by simply indicating the instantaneous offset from the nominal value to be made, and the corresponding element of the data signal is identified thereby. If the data signal is modulated by a noise-like signal, demodulation with the same noise-like signal gives the original information signal. (Such demodulation by a noise signal is not necessary if the original information signal serves as a basis for the clock phase shift.)

More typically, the channel to the remote computer introduces temporal distortion. One example is the Internet, where different packets follow different routing between the originating computer and the receiving computer. In such environments, thousands of edge changes corresponding to a given bit are averaged or otherwise processed to determine whether the edge is leading or lagging, thereby raising this bit. Decoding may need to be performed with some degree of statistical guarantee. Thus, in environments with highly variable transmission call times (eg, the cited Internet example), 100-bit messages can be collected at edge changes on the order of one million to ensure reliable message recovery. And may require processing. However, with increasing data rates on the Internet, this may require only a few seconds of data.

In an illustrative embodiment, the master clock of the computer is configured as a phase modulator and driven by the data signal in a phase locked loop (PLL).
) Controlled by the circuit. The decoding is performed by another PLL configured as a phase demodulator. (The two PLLs do not need to share the same time axis because decoding can proceed by examining the deviation of the instantaneous edge change from the average clock period.) PLL phase modulation such as these The design of the demodulator and demodulator
It is within the ordinary ability of those skilled in the art.

[0149] Given the many possible embodiments to which the principles of the present invention may be directed, the detailed embodiments are illustrative only and should not be taken to limit the scope of the present invention.
Rather, all such embodiments are claimed as the invention, which comes within the scope and spirit of the following claims and their equivalents.

[Brief description of the drawings]

1A and 1B show prior art techniques for achieving a grayscale effect using line drawings.

FIG. 2 illustrates a virtual array of grid points that can be overlaid on an image according to one embodiment of the present invention.

FIG. 3 shows a virtual array of regions that can be overlaid on an image according to the embodiment of FIG.

FIG. 4 shows an excerpt from FIG. 3 with a line from a passing line drawing image.

FIG. 5 illustrates performing watermark encoding with varying line widths in FIG. 3, according to one embodiment of the present invention.

FIG. 6 illustrates performing watermark encoding by changing the position of the line in FIG. 3 according to another embodiment of the present invention.

FIG. 7 is a block diagram of a photocopier according to another embodiment of the present invention.

FIG. 8 illustrates a portion of an automatic teller machine using the principles of the present invention.

FIG. 9 illustrates a portion of an apparatus (eg, a photocopier, scanner, or printer) that uses the principles of the present invention.

FIG. 10 illustrates an apparatus (eg, a photocopier,
(Scanner or printer).

──────────────────────────────────────────────────続 き Continuation of front page (81) Designated country EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE ), OA (BF, BJ, CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG), AP (GH, GM, KE, LS, MW, SD, SL, SZ, UG, ZW), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), AE, AL, AM, AT, AU, AZ, BA, BB, BG, BR , BY, CA, CH, CN, CU, CZ, DE, DK, EE, ES, FI, GB, GD, GE, GH, GM, HR, HU, ID, IL, IN, IS , JP, KE, KG, KP, KR, KZ, LC, LK, LR, LS, LT, LU, LV, MD, MG, MK, MN, MW, MX, NO, NZ, PL, PT, RO, RU, SD, SE, SG, SI, SK, SL, TJ, TM, TR, TT, UA, UG, US, UZ, VN, YU, ZA, ZWF terms (reference) 3E041 AA03 BA09 BA13 BB02 BC03 CB03 DB01 5B057 AA11 CA01 CA08 CA16 CA19 CB01 CB08 CB19 CC01 CE08 CG05 DA02 DB06 DB09 5C076 AA14 BA02 CA06 5C077 LL14 MP01 MP08 PP23 TT06

Claims (38)

[Claims] 1. A method of increasing the security of a printed bill, the method comprising marking machine-readable, generally imperceptible data on the face of the bill, wherein the bill is to be marked. A method comprising: selecting digital data; and slightly changing a distribution of ink on a face of the banknote to encode the digital data. 2. The method of claim 1, wherein said digital data comprises a plurality of bits. 3. The method of claim 1 wherein, instead of marking only one localized area on the bill, all of the plurality of bits are redundantly encoded across the bill. how to. 4. The method according to claim 1, wherein said encoding uses a code signal. 5. The method of claim 1, wherein said encoding uses a discrete cosine transform. 6. The method according to claim 1, comprising encoding two different digital watermarks. 7. The method according to claim 6, wherein the two different digital watermarks are of different robustness. 8. The method according to claim 6, wherein said two watermarks are encoded by different code signals. 9. The method of claim 1, further comprising providing a hologram on the bill. 10. A method of encoding multi-bit digital data on a banknote to facilitate subsequent machine recognition of the banknote, comprising: generating a pattern of data encoding the multi-bit digital data; A method comprising: receiving an artwork; adjusting the initial banknote artwork according to the data; and printing a banknote corresponding to the adjusted banknote artwork. 11. The method according to claim 10, wherein the initial banknote artwork comprises a line drawing, the method changing the positions of the lines making up the artwork slightly and encoding the multi-bit data. A method comprising: 12. The method of claim 11, wherein the position of at least one line is slightly changed in a first direction in a first region along the at least one line, and the position is along the line. Making a slight change in a second direction in the second region that is different from the first direction. 13. The method of claim 10, wherein the banknote artwork comprises a line drawing, the method comprising slightly varying the thickness of the lines comprising the artwork and encoding the multi-bit data. A method comprising: 14. The method according to claim 13, comprising, for at least one line, increasing its thickness in a first region and decreasing its thickness in a second region. how to. 15. The method of claim 10, further comprising: defining an array of regions that span at least a portion of the initial banknote artwork; and varying the artwork in a plurality of the regions to each other. Increasing or decreasing the target brightness. 16. The method of claim 15, wherein said defining comprises overlaying a virtual grid over said portion. 17. The method of claim 10, wherein the original artwork is:
A method comprising adjusting by inserting one or more new lines therein. 18. A method for preventing banknote duplication, comprising: encoding a plurality of bits of digital data on the banknote; providing sampled image data corresponding to the banknote; analyzing the image data; Detecting the digital data at wherein the detection means that the image data appears to correspond to a banknote;
Interfering in response to the detection of the digital data. 19. The method according to claim 18, comprising encoding the bill with the digital data across its surface, the method comprising: erasing localized marking features of the bill; A method characterized in that it cannot be defeated by reinserting such markings. 20. The method of claim 18, further comprising: analyzing the image data to detect a visible structural feature of a banknote; and interfering in response to detecting the visible structure. A method characterized in that detection of either structure or digital data serves to trigger said interference. 21. The method of claim 18, wherein said interfering comprises preventing duplication of said image data. 22. The method of claim 18, wherein said interfering comprises sending an eye sage to a remote location to indicate processing of a bill. 23. The method of claim 18, wherein said interfering comprises inserting legal tracking data into said image data. 24. The method of claim 22, wherein the interference is a first and a second.
A method comprising inserting a multi-bit digital watermark into said image data, one of said watermarks serving to carry data identifying a device involved in an attempt to duplicate said bill. 25. The method of claim 24, wherein the first and second watermarks are of different robustness. 26. The method according to claim 24, wherein the first and second watermarks are encoded according to different code signals. 27. A method for preventing banknote duplication, comprising: inspecting, at a first site, suspicious image data corresponding to a sampled optical brightness of an object; and, at the first site, the suspicious image. Determining whether or not corresponds to a bill, and if so, contacting a second site and reporting detection of data related to the bill. 28. A method for marking printer output from a personal computer system and enabling subsequent recognition, said system comprising at least a personal computer and a printer, said method comprising: Receiving the printer data, encoding a digital watermark comprising a plurality of bits of digital data that is machine-readable and generally imperceptible, the digital watermark comprising the printer data in the printer data. Helping to mark as being printed by the printer, the data printed by the printer having the digital watermark, enabling the printer to print the document recognizable. Features to And how. 29. An apparatus for checking bills, comprising: an optical scanner for generating image data corresponding to a document input to the apparatus; detecting watermark data in the image data, wherein the watermark data includes a plurality of digital bits. A processor, wherein the detection of such watermark data indicates that the document is a banknote; if the processor detects watermark data in the image data, responding in a first manner, wherein the processor A control unit responsive in a second different manner if no watermark data is detected in the image data. 30. The apparatus according to claim 29, wherein the processor has an input supplied with a code signal, the code signal being necessary to distinguish the watermark data from the image data. An apparatus characterized by the above. 31. The apparatus according to claim 29, wherein the processor analyzes the image data for visible structural features of a bill, and wherein the control unit determines that both the watermark data and the visible structure are in the image data. Apparatus for responding in the first method only if detected by the processor. 32. The automatic teller machine according to claim 31, wherein the machine performs both receipt and payment of bills from a user. 33. An apparatus for detecting an attempted counterfeit of a banknote, comprising: a banknote having a line drawing on a surface thereof; a scanner for scanning the banknote; and binary data encoded in the banknote line drawing and scanned by the scanner. And a steganographic watermark detector for decrypting the banknote, wherein the presence of the decrypted binary data indicates an attempted counterfeit of the banknote. 34. The apparatus according to claim 33, further comprising a control unit responsive to the decoded binary data to initiate insertion of tracking data into output generated by the apparatus. Equipment to do. 35. The apparatus according to claim 34, wherein the tracking data includes data relating to a date of duplication. 36. The apparatus of claim 34, wherein the tracking data comprises a steganographic watermark. 37. The apparatus according to claim 34, wherein there is a delay in the decryption so that when scanning bills, the apparatus first produces an output without the tracking data inserted. And wherein the subsequent output includes the tracking data. 38. An apparatus for processing cash, comprising: an input unit for receiving a plurality of bills; a feed mechanism for transporting the bills from the input unit; and an optic for generating digital image data corresponding to the bills transported by the feed mechanism. A detector for detecting steganographically encoded multi-bit digital data in the artwork of at least some of the bill, and a control unit responsive to the detection of the multi-bit digital data. And equipment.